1 //===- Writer.cpp ---------------------------------------------------------===// 2 // 3 // The LLVM Linker 4 // 5 // This file is distributed under the University of Illinois Open Source 6 // License. See LICENSE.TXT for details. 7 // 8 //===----------------------------------------------------------------------===// 9 10 #include "Writer.h" 11 #include "AArch64ErrataFix.h" 12 #include "Config.h" 13 #include "Filesystem.h" 14 #include "LinkerScript.h" 15 #include "MapFile.h" 16 #include "OutputSections.h" 17 #include "Relocations.h" 18 #include "Strings.h" 19 #include "SymbolTable.h" 20 #include "Symbols.h" 21 #include "SyntheticSections.h" 22 #include "Target.h" 23 #include "lld/Common/Memory.h" 24 #include "lld/Common/Threads.h" 25 #include "llvm/ADT/StringMap.h" 26 #include "llvm/ADT/StringSwitch.h" 27 #include <climits> 28 29 using namespace llvm; 30 using namespace llvm::ELF; 31 using namespace llvm::object; 32 using namespace llvm::support; 33 using namespace llvm::support::endian; 34 35 using namespace lld; 36 using namespace lld::elf; 37 38 namespace { 39 // The writer writes a SymbolTable result to a file. 40 template <class ELFT> class Writer { 41 public: 42 Writer() : Buffer(errorHandler().OutputBuffer) {} 43 typedef typename ELFT::Shdr Elf_Shdr; 44 typedef typename ELFT::Ehdr Elf_Ehdr; 45 typedef typename ELFT::Phdr Elf_Phdr; 46 47 void run(); 48 49 private: 50 void copyLocalSymbols(); 51 void addSectionSymbols(); 52 void forEachRelSec(std::function<void(InputSectionBase &)> Fn); 53 void sortSections(); 54 void resolveShfLinkOrder(); 55 void sortInputSections(); 56 void finalizeSections(); 57 void addPredefinedSections(); 58 void setReservedSymbolSections(); 59 60 std::vector<PhdrEntry *> createPhdrs(); 61 void removeEmptyPTLoad(); 62 void addPtArmExid(std::vector<PhdrEntry *> &Phdrs); 63 void assignFileOffsets(); 64 void assignFileOffsetsBinary(); 65 void setPhdrs(); 66 void fixSectionAlignments(); 67 void openFile(); 68 void writeTrapInstr(); 69 void writeHeader(); 70 void writeSections(); 71 void writeSectionsBinary(); 72 void writeBuildId(); 73 74 std::unique_ptr<FileOutputBuffer> &Buffer; 75 76 void addRelIpltSymbols(); 77 void addStartEndSymbols(); 78 void addStartStopSymbols(OutputSection *Sec); 79 uint64_t getEntryAddr(); 80 81 std::vector<PhdrEntry *> Phdrs; 82 83 uint64_t FileSize; 84 uint64_t SectionHeaderOff; 85 86 bool HasGotBaseSym = false; 87 }; 88 } // anonymous namespace 89 90 StringRef elf::getOutputSectionName(InputSectionBase *S) { 91 // ".zdebug_" is a prefix for ZLIB-compressed sections. 92 // Because we decompressed input sections, we want to remove 'z'. 93 if (S->Name.startswith(".zdebug_")) 94 return Saver.save("." + S->Name.substr(2)); 95 96 if (Config->Relocatable) 97 return S->Name; 98 99 // This is for --emit-relocs. If .text.foo is emitted as .text.bar, we want 100 // to emit .rela.text.foo as .rela.text.bar for consistency (this is not 101 // technically required, but not doing it is odd). This code guarantees that. 102 if ((S->Type == SHT_REL || S->Type == SHT_RELA) && 103 !isa<SyntheticSection>(S)) { 104 OutputSection *Out = 105 cast<InputSection>(S)->getRelocatedSection()->getOutputSection(); 106 if (S->Type == SHT_RELA) 107 return Saver.save(".rela" + Out->Name); 108 return Saver.save(".rel" + Out->Name); 109 } 110 111 for (StringRef V : 112 {".text.", ".rodata.", ".data.rel.ro.", ".data.", ".bss.rel.ro.", 113 ".bss.", ".init_array.", ".fini_array.", ".ctors.", ".dtors.", ".tbss.", 114 ".gcc_except_table.", ".tdata.", ".ARM.exidx.", ".ARM.extab."}) { 115 StringRef Prefix = V.drop_back(); 116 if (S->Name.startswith(V) || S->Name == Prefix) 117 return Prefix; 118 } 119 120 // CommonSection is identified as "COMMON" in linker scripts. 121 // By default, it should go to .bss section. 122 if (S->Name == "COMMON") 123 return ".bss"; 124 125 return S->Name; 126 } 127 128 static bool needsInterpSection() { 129 return !SharedFiles.empty() && !Config->DynamicLinker.empty() && 130 Script->needsInterpSection(); 131 } 132 133 template <class ELFT> void elf::writeResult() { Writer<ELFT>().run(); } 134 135 template <class ELFT> void Writer<ELFT>::removeEmptyPTLoad() { 136 llvm::erase_if(Phdrs, [&](const PhdrEntry *P) { 137 if (P->p_type != PT_LOAD) 138 return false; 139 if (!P->FirstSec) 140 return true; 141 uint64_t Size = P->LastSec->Addr + P->LastSec->Size - P->FirstSec->Addr; 142 return Size == 0; 143 }); 144 } 145 146 template <class ELFT> static void combineEhFrameSections() { 147 for (InputSectionBase *&S : InputSections) { 148 EhInputSection *ES = dyn_cast<EhInputSection>(S); 149 if (!ES || !ES->Live) 150 continue; 151 152 InX::EhFrame->addSection<ELFT>(ES); 153 S = nullptr; 154 } 155 156 std::vector<InputSectionBase *> &V = InputSections; 157 V.erase(std::remove(V.begin(), V.end(), nullptr), V.end()); 158 } 159 160 template <class ELFT> 161 static Defined *addOptionalRegular(StringRef Name, SectionBase *Sec, 162 uint64_t Val, uint8_t StOther = STV_HIDDEN, 163 uint8_t Binding = STB_GLOBAL) { 164 Symbol *S = Symtab->find(Name); 165 if (!S || S->isDefined()) 166 return nullptr; 167 Symbol *Sym = Symtab->addRegular<ELFT>(Name, StOther, STT_NOTYPE, Val, 168 /*Size=*/0, Binding, Sec, 169 /*File=*/nullptr); 170 return cast<Defined>(Sym); 171 } 172 173 // The linker is expected to define some symbols depending on 174 // the linking result. This function defines such symbols. 175 template <class ELFT> void elf::addReservedSymbols() { 176 if (Config->EMachine == EM_MIPS) { 177 // Define _gp for MIPS. st_value of _gp symbol will be updated by Writer 178 // so that it points to an absolute address which by default is relative 179 // to GOT. Default offset is 0x7ff0. 180 // See "Global Data Symbols" in Chapter 6 in the following document: 181 // ftp://www.linux-mips.org/pub/linux/mips/doc/ABI/mipsabi.pdf 182 ElfSym::MipsGp = Symtab->addAbsolute<ELFT>("_gp", STV_HIDDEN, STB_GLOBAL); 183 184 // On MIPS O32 ABI, _gp_disp is a magic symbol designates offset between 185 // start of function and 'gp' pointer into GOT. 186 if (Symtab->find("_gp_disp")) 187 ElfSym::MipsGpDisp = 188 Symtab->addAbsolute<ELFT>("_gp_disp", STV_HIDDEN, STB_GLOBAL); 189 190 // The __gnu_local_gp is a magic symbol equal to the current value of 'gp' 191 // pointer. This symbol is used in the code generated by .cpload pseudo-op 192 // in case of using -mno-shared option. 193 // https://sourceware.org/ml/binutils/2004-12/msg00094.html 194 if (Symtab->find("__gnu_local_gp")) 195 ElfSym::MipsLocalGp = 196 Symtab->addAbsolute<ELFT>("__gnu_local_gp", STV_HIDDEN, STB_GLOBAL); 197 } 198 199 ElfSym::GlobalOffsetTable = addOptionalRegular<ELFT>( 200 "_GLOBAL_OFFSET_TABLE_", Out::ElfHeader, Target->GotBaseSymOff); 201 202 // __ehdr_start is the location of ELF file headers. Note that we define 203 // this symbol unconditionally even when using a linker script, which 204 // differs from the behavior implemented by GNU linker which only define 205 // this symbol if ELF headers are in the memory mapped segment. 206 // __executable_start is not documented, but the expectation of at 207 // least the android libc is that it points to the elf header too. 208 // __dso_handle symbol is passed to cxa_finalize as a marker to identify 209 // each DSO. The address of the symbol doesn't matter as long as they are 210 // different in different DSOs, so we chose the start address of the DSO. 211 for (const char *Name : 212 {"__ehdr_start", "__executable_start", "__dso_handle"}) 213 addOptionalRegular<ELFT>(Name, Out::ElfHeader, 0, STV_HIDDEN); 214 215 // If linker script do layout we do not need to create any standart symbols. 216 if (Script->HasSectionsCommand) 217 return; 218 219 auto Add = [](StringRef S, int64_t Pos) { 220 return addOptionalRegular<ELFT>(S, Out::ElfHeader, Pos, STV_DEFAULT); 221 }; 222 223 ElfSym::Bss = Add("__bss_start", 0); 224 ElfSym::End1 = Add("end", -1); 225 ElfSym::End2 = Add("_end", -1); 226 ElfSym::Etext1 = Add("etext", -1); 227 ElfSym::Etext2 = Add("_etext", -1); 228 ElfSym::Edata1 = Add("edata", -1); 229 ElfSym::Edata2 = Add("_edata", -1); 230 } 231 232 static OutputSection *findSection(StringRef Name) { 233 for (BaseCommand *Base : Script->SectionCommands) 234 if (auto *Sec = dyn_cast<OutputSection>(Base)) 235 if (Sec->Name == Name) 236 return Sec; 237 return nullptr; 238 } 239 240 // Initialize Out members. 241 template <class ELFT> static void createSyntheticSections() { 242 // Initialize all pointers with NULL. This is needed because 243 // you can call lld::elf::main more than once as a library. 244 memset(&Out::First, 0, sizeof(Out)); 245 246 auto Add = [](InputSectionBase *Sec) { InputSections.push_back(Sec); }; 247 248 InX::DynStrTab = make<StringTableSection>(".dynstr", true); 249 InX::Dynamic = make<DynamicSection<ELFT>>(); 250 if (Config->AndroidPackDynRelocs) { 251 InX::RelaDyn = make<AndroidPackedRelocationSection<ELFT>>( 252 Config->IsRela ? ".rela.dyn" : ".rel.dyn"); 253 } else { 254 InX::RelaDyn = make<RelocationSection<ELFT>>( 255 Config->IsRela ? ".rela.dyn" : ".rel.dyn", Config->ZCombreloc); 256 } 257 InX::ShStrTab = make<StringTableSection>(".shstrtab", false); 258 259 Out::ProgramHeaders = make<OutputSection>("", 0, SHF_ALLOC); 260 Out::ProgramHeaders->Alignment = Config->Wordsize; 261 262 if (needsInterpSection()) { 263 InX::Interp = createInterpSection(); 264 Add(InX::Interp); 265 } else { 266 InX::Interp = nullptr; 267 } 268 269 if (Config->Strip != StripPolicy::All) { 270 InX::StrTab = make<StringTableSection>(".strtab", false); 271 InX::SymTab = make<SymbolTableSection<ELFT>>(*InX::StrTab); 272 } 273 274 if (Config->BuildId != BuildIdKind::None) { 275 InX::BuildId = make<BuildIdSection>(); 276 Add(InX::BuildId); 277 } 278 279 InX::Bss = make<BssSection>(".bss", 0, 1); 280 Add(InX::Bss); 281 282 // If there is a SECTIONS command and a .data.rel.ro section name use name 283 // .data.rel.ro.bss so that we match in the .data.rel.ro output section. 284 // This makes sure our relro is contiguous. 285 bool HasDataRelRo = 286 Script->HasSectionsCommand && findSection(".data.rel.ro"); 287 InX::BssRelRo = make<BssSection>( 288 HasDataRelRo ? ".data.rel.ro.bss" : ".bss.rel.ro", 0, 1); 289 Add(InX::BssRelRo); 290 291 // Add MIPS-specific sections. 292 if (Config->EMachine == EM_MIPS) { 293 if (!Config->Shared && Config->HasDynSymTab) { 294 InX::MipsRldMap = make<MipsRldMapSection>(); 295 Add(InX::MipsRldMap); 296 } 297 if (auto *Sec = MipsAbiFlagsSection<ELFT>::create()) 298 Add(Sec); 299 if (auto *Sec = MipsOptionsSection<ELFT>::create()) 300 Add(Sec); 301 if (auto *Sec = MipsReginfoSection<ELFT>::create()) 302 Add(Sec); 303 } 304 305 if (Config->HasDynSymTab) { 306 InX::DynSymTab = make<SymbolTableSection<ELFT>>(*InX::DynStrTab); 307 Add(InX::DynSymTab); 308 309 In<ELFT>::VerSym = make<VersionTableSection<ELFT>>(); 310 Add(In<ELFT>::VerSym); 311 312 if (!Config->VersionDefinitions.empty()) { 313 In<ELFT>::VerDef = make<VersionDefinitionSection<ELFT>>(); 314 Add(In<ELFT>::VerDef); 315 } 316 317 In<ELFT>::VerNeed = make<VersionNeedSection<ELFT>>(); 318 Add(In<ELFT>::VerNeed); 319 320 if (Config->GnuHash) { 321 InX::GnuHashTab = make<GnuHashTableSection>(); 322 Add(InX::GnuHashTab); 323 } 324 325 if (Config->SysvHash) { 326 InX::HashTab = make<HashTableSection>(); 327 Add(InX::HashTab); 328 } 329 330 Add(InX::Dynamic); 331 Add(InX::DynStrTab); 332 Add(InX::RelaDyn); 333 } 334 335 // Add .got. MIPS' .got is so different from the other archs, 336 // it has its own class. 337 if (Config->EMachine == EM_MIPS) { 338 InX::MipsGot = make<MipsGotSection>(); 339 Add(InX::MipsGot); 340 } else { 341 InX::Got = make<GotSection>(); 342 Add(InX::Got); 343 } 344 345 InX::GotPlt = make<GotPltSection>(); 346 Add(InX::GotPlt); 347 InX::IgotPlt = make<IgotPltSection>(); 348 Add(InX::IgotPlt); 349 350 if (Config->GdbIndex) { 351 InX::GdbIndex = createGdbIndex<ELFT>(); 352 Add(InX::GdbIndex); 353 } 354 355 // We always need to add rel[a].plt to output if it has entries. 356 // Even for static linking it can contain R_[*]_IRELATIVE relocations. 357 InX::RelaPlt = make<RelocationSection<ELFT>>( 358 Config->IsRela ? ".rela.plt" : ".rel.plt", false /*Sort*/); 359 Add(InX::RelaPlt); 360 361 // The RelaIplt immediately follows .rel.plt (.rel.dyn for ARM) to ensure 362 // that the IRelative relocations are processed last by the dynamic loader. 363 // We cannot place the iplt section in .rel.dyn when Android relocation 364 // packing is enabled because that would cause a section type mismatch. 365 // However, because the Android dynamic loader reads .rel.plt after .rel.dyn, 366 // we can get the desired behaviour by placing the iplt section in .rel.plt. 367 InX::RelaIplt = make<RelocationSection<ELFT>>( 368 (Config->EMachine == EM_ARM && !Config->AndroidPackDynRelocs) 369 ? ".rel.dyn" 370 : InX::RelaPlt->Name, 371 false /*Sort*/); 372 Add(InX::RelaIplt); 373 374 InX::Plt = make<PltSection>(Target->PltHeaderSize); 375 Add(InX::Plt); 376 InX::Iplt = make<PltSection>(0); 377 Add(InX::Iplt); 378 379 if (!Config->Relocatable) { 380 if (Config->EhFrameHdr) { 381 InX::EhFrameHdr = make<EhFrameHeader>(); 382 Add(InX::EhFrameHdr); 383 } 384 InX::EhFrame = make<EhFrameSection>(); 385 Add(InX::EhFrame); 386 } 387 388 if (InX::SymTab) 389 Add(InX::SymTab); 390 Add(InX::ShStrTab); 391 if (InX::StrTab) 392 Add(InX::StrTab); 393 } 394 395 // The main function of the writer. 396 template <class ELFT> void Writer<ELFT>::run() { 397 // Create linker-synthesized sections such as .got or .plt. 398 // Such sections are of type input section. 399 createSyntheticSections<ELFT>(); 400 401 if (!Config->Relocatable) 402 combineEhFrameSections<ELFT>(); 403 404 // We want to process linker script commands. When SECTIONS command 405 // is given we let it create sections. 406 Script->processSectionCommands(); 407 408 // Linker scripts controls how input sections are assigned to output sections. 409 // Input sections that were not handled by scripts are called "orphans", and 410 // they are assigned to output sections by the default rule. Process that. 411 Script->addOrphanSections(); 412 413 if (Config->Discard != DiscardPolicy::All) 414 copyLocalSymbols(); 415 416 if (Config->CopyRelocs) 417 addSectionSymbols(); 418 419 // Now that we have a complete set of output sections. This function 420 // completes section contents. For example, we need to add strings 421 // to the string table, and add entries to .got and .plt. 422 // finalizeSections does that. 423 finalizeSections(); 424 if (errorCount()) 425 return; 426 427 // If -compressed-debug-sections is specified, we need to compress 428 // .debug_* sections. Do it right now because it changes the size of 429 // output sections. 430 parallelForEach(OutputSections, 431 [](OutputSection *Sec) { Sec->maybeCompress<ELFT>(); }); 432 433 Script->assignAddresses(); 434 Script->allocateHeaders(Phdrs); 435 436 // Remove empty PT_LOAD to avoid causing the dynamic linker to try to mmap a 437 // 0 sized region. This has to be done late since only after assignAddresses 438 // we know the size of the sections. 439 removeEmptyPTLoad(); 440 441 if (!Config->OFormatBinary) 442 assignFileOffsets(); 443 else 444 assignFileOffsetsBinary(); 445 446 setPhdrs(); 447 448 if (Config->Relocatable) { 449 for (OutputSection *Sec : OutputSections) 450 Sec->Addr = 0; 451 } 452 453 // It does not make sense try to open the file if we have error already. 454 if (errorCount()) 455 return; 456 // Write the result down to a file. 457 openFile(); 458 if (errorCount()) 459 return; 460 461 if (!Config->OFormatBinary) { 462 writeTrapInstr(); 463 writeHeader(); 464 writeSections(); 465 } else { 466 writeSectionsBinary(); 467 } 468 469 // Backfill .note.gnu.build-id section content. This is done at last 470 // because the content is usually a hash value of the entire output file. 471 writeBuildId(); 472 if (errorCount()) 473 return; 474 475 // Handle -Map option. 476 writeMapFile(); 477 if (errorCount()) 478 return; 479 480 if (auto E = Buffer->commit()) 481 error("failed to write to the output file: " + toString(std::move(E))); 482 } 483 484 static bool shouldKeepInSymtab(SectionBase *Sec, StringRef SymName, 485 const Symbol &B) { 486 if (B.isFile() || B.isSection()) 487 return false; 488 489 // If sym references a section in a discarded group, don't keep it. 490 if (Sec == &InputSection::Discarded) 491 return false; 492 493 if (Config->Discard == DiscardPolicy::None) 494 return true; 495 496 // In ELF assembly .L symbols are normally discarded by the assembler. 497 // If the assembler fails to do so, the linker discards them if 498 // * --discard-locals is used. 499 // * The symbol is in a SHF_MERGE section, which is normally the reason for 500 // the assembler keeping the .L symbol. 501 if (!SymName.startswith(".L") && !SymName.empty()) 502 return true; 503 504 if (Config->Discard == DiscardPolicy::Locals) 505 return false; 506 507 return !Sec || !(Sec->Flags & SHF_MERGE); 508 } 509 510 static bool includeInSymtab(const Symbol &B) { 511 if (!B.isLocal() && !B.IsUsedInRegularObj) 512 return false; 513 514 if (auto *D = dyn_cast<Defined>(&B)) { 515 // Always include absolute symbols. 516 SectionBase *Sec = D->Section; 517 if (!Sec) 518 return true; 519 Sec = Sec->Repl; 520 // Exclude symbols pointing to garbage-collected sections. 521 if (isa<InputSectionBase>(Sec) && !Sec->Live) 522 return false; 523 if (auto *S = dyn_cast<MergeInputSection>(Sec)) 524 if (!S->getSectionPiece(D->Value)->Live) 525 return false; 526 return true; 527 } 528 return B.Used; 529 } 530 531 // Local symbols are not in the linker's symbol table. This function scans 532 // each object file's symbol table to copy local symbols to the output. 533 template <class ELFT> void Writer<ELFT>::copyLocalSymbols() { 534 if (!InX::SymTab) 535 return; 536 for (InputFile *File : ObjectFiles) { 537 ObjFile<ELFT> *F = cast<ObjFile<ELFT>>(File); 538 for (Symbol *B : F->getLocalSymbols()) { 539 if (!B->isLocal()) 540 fatal(toString(F) + 541 ": broken object: getLocalSymbols returns a non-local symbol"); 542 auto *DR = dyn_cast<Defined>(B); 543 544 // No reason to keep local undefined symbol in symtab. 545 if (!DR) 546 continue; 547 if (!includeInSymtab(*B)) 548 continue; 549 550 SectionBase *Sec = DR->Section; 551 if (!shouldKeepInSymtab(Sec, B->getName(), *B)) 552 continue; 553 InX::SymTab->addSymbol(B); 554 } 555 } 556 } 557 558 template <class ELFT> void Writer<ELFT>::addSectionSymbols() { 559 // Create a section symbol for each output section so that we can represent 560 // relocations that point to the section. If we know that no relocation is 561 // referring to a section (that happens if the section is a synthetic one), we 562 // don't create a section symbol for that section. 563 for (BaseCommand *Base : Script->SectionCommands) { 564 auto *Sec = dyn_cast<OutputSection>(Base); 565 if (!Sec) 566 continue; 567 auto I = llvm::find_if(Sec->SectionCommands, [](BaseCommand *Base) { 568 if (auto *ISD = dyn_cast<InputSectionDescription>(Base)) 569 return !ISD->Sections.empty(); 570 return false; 571 }); 572 if (I == Sec->SectionCommands.end()) 573 continue; 574 InputSection *IS = cast<InputSectionDescription>(*I)->Sections[0]; 575 576 // Relocations are not using REL[A] section symbols. 577 if (IS->Type == SHT_REL || IS->Type == SHT_RELA) 578 continue; 579 580 // Unlike other synthetic sections, mergeable output sections contain data 581 // copied from input sections, and there may be a relocation pointing to its 582 // contents if -r or -emit-reloc are given. 583 if (isa<SyntheticSection>(IS) && !(IS->Flags & SHF_MERGE)) 584 continue; 585 586 auto *Sym = 587 make<Defined>(IS->File, "", STB_LOCAL, /*StOther=*/0, STT_SECTION, 588 /*Value=*/0, /*Size=*/0, IS); 589 InX::SymTab->addSymbol(Sym); 590 } 591 } 592 593 // Today's loaders have a feature to make segments read-only after 594 // processing dynamic relocations to enhance security. PT_GNU_RELRO 595 // is defined for that. 596 // 597 // This function returns true if a section needs to be put into a 598 // PT_GNU_RELRO segment. 599 static bool isRelroSection(const OutputSection *Sec) { 600 if (!Config->ZRelro) 601 return false; 602 603 uint64_t Flags = Sec->Flags; 604 605 // Non-allocatable or non-writable sections don't need RELRO because 606 // they are not writable or not even mapped to memory in the first place. 607 // RELRO is for sections that are essentially read-only but need to 608 // be writable only at process startup to allow dynamic linker to 609 // apply relocations. 610 if (!(Flags & SHF_ALLOC) || !(Flags & SHF_WRITE)) 611 return false; 612 613 // Once initialized, TLS data segments are used as data templates 614 // for a thread-local storage. For each new thread, runtime 615 // allocates memory for a TLS and copy templates there. No thread 616 // are supposed to use templates directly. Thus, it can be in RELRO. 617 if (Flags & SHF_TLS) 618 return true; 619 620 // .init_array, .preinit_array and .fini_array contain pointers to 621 // functions that are executed on process startup or exit. These 622 // pointers are set by the static linker, and they are not expected 623 // to change at runtime. But if you are an attacker, you could do 624 // interesting things by manipulating pointers in .fini_array, for 625 // example. So they are put into RELRO. 626 uint32_t Type = Sec->Type; 627 if (Type == SHT_INIT_ARRAY || Type == SHT_FINI_ARRAY || 628 Type == SHT_PREINIT_ARRAY) 629 return true; 630 631 // .got contains pointers to external symbols. They are resolved by 632 // the dynamic linker when a module is loaded into memory, and after 633 // that they are not expected to change. So, it can be in RELRO. 634 if (InX::Got && Sec == InX::Got->getParent()) 635 return true; 636 637 // .got.plt contains pointers to external function symbols. They are 638 // by default resolved lazily, so we usually cannot put it into RELRO. 639 // However, if "-z now" is given, the lazy symbol resolution is 640 // disabled, which enables us to put it into RELRO. 641 if (Sec == InX::GotPlt->getParent()) 642 return Config->ZNow; 643 644 // .dynamic section contains data for the dynamic linker, and 645 // there's no need to write to it at runtime, so it's better to put 646 // it into RELRO. 647 if (Sec == InX::Dynamic->getParent()) 648 return true; 649 650 // Sections with some special names are put into RELRO. This is a 651 // bit unfortunate because section names shouldn't be significant in 652 // ELF in spirit. But in reality many linker features depend on 653 // magic section names. 654 StringRef S = Sec->Name; 655 return S == ".data.rel.ro" || S == ".bss.rel.ro" || S == ".ctors" || 656 S == ".dtors" || S == ".jcr" || S == ".eh_frame" || 657 S == ".openbsd.randomdata"; 658 } 659 660 // We compute a rank for each section. The rank indicates where the 661 // section should be placed in the file. Instead of using simple 662 // numbers (0,1,2...), we use a series of flags. One for each decision 663 // point when placing the section. 664 // Using flags has two key properties: 665 // * It is easy to check if a give branch was taken. 666 // * It is easy two see how similar two ranks are (see getRankProximity). 667 enum RankFlags { 668 RF_NOT_ADDR_SET = 1 << 16, 669 RF_NOT_INTERP = 1 << 15, 670 RF_NOT_ALLOC = 1 << 14, 671 RF_WRITE = 1 << 13, 672 RF_EXEC_WRITE = 1 << 12, 673 RF_EXEC = 1 << 11, 674 RF_NON_TLS_BSS = 1 << 10, 675 RF_NON_TLS_BSS_RO = 1 << 9, 676 RF_NOT_TLS = 1 << 8, 677 RF_BSS = 1 << 7, 678 RF_PPC_NOT_TOCBSS = 1 << 6, 679 RF_PPC_OPD = 1 << 5, 680 RF_PPC_TOCL = 1 << 4, 681 RF_PPC_TOC = 1 << 3, 682 RF_PPC_BRANCH_LT = 1 << 2, 683 RF_MIPS_GPREL = 1 << 1, 684 RF_MIPS_NOT_GOT = 1 << 0 685 }; 686 687 static unsigned getSectionRank(const OutputSection *Sec) { 688 unsigned Rank = 0; 689 690 // We want to put section specified by -T option first, so we 691 // can start assigning VA starting from them later. 692 if (Config->SectionStartMap.count(Sec->Name)) 693 return Rank; 694 Rank |= RF_NOT_ADDR_SET; 695 696 // Put .interp first because some loaders want to see that section 697 // on the first page of the executable file when loaded into memory. 698 if (Sec->Name == ".interp") 699 return Rank; 700 Rank |= RF_NOT_INTERP; 701 702 // Allocatable sections go first to reduce the total PT_LOAD size and 703 // so debug info doesn't change addresses in actual code. 704 if (!(Sec->Flags & SHF_ALLOC)) 705 return Rank | RF_NOT_ALLOC; 706 707 // Sort sections based on their access permission in the following 708 // order: R, RX, RWX, RW. This order is based on the following 709 // considerations: 710 // * Read-only sections come first such that they go in the 711 // PT_LOAD covering the program headers at the start of the file. 712 // * Read-only, executable sections come next, unless the 713 // -no-rosegment option is used. 714 // * Writable, executable sections follow such that .plt on 715 // architectures where it needs to be writable will be placed 716 // between .text and .data. 717 // * Writable sections come last, such that .bss lands at the very 718 // end of the last PT_LOAD. 719 bool IsExec = Sec->Flags & SHF_EXECINSTR; 720 bool IsWrite = Sec->Flags & SHF_WRITE; 721 722 if (IsExec) { 723 if (IsWrite) 724 Rank |= RF_EXEC_WRITE; 725 else if (!Config->SingleRoRx) 726 Rank |= RF_EXEC; 727 } else { 728 if (IsWrite) 729 Rank |= RF_WRITE; 730 } 731 732 // If we got here we know that both A and B are in the same PT_LOAD. 733 734 bool IsTls = Sec->Flags & SHF_TLS; 735 bool IsNoBits = Sec->Type == SHT_NOBITS; 736 737 // The first requirement we have is to put (non-TLS) nobits sections last. The 738 // reason is that the only thing the dynamic linker will see about them is a 739 // p_memsz that is larger than p_filesz. Seeing that it zeros the end of the 740 // PT_LOAD, so that has to correspond to the nobits sections. 741 bool IsNonTlsNoBits = IsNoBits && !IsTls; 742 if (IsNonTlsNoBits) 743 Rank |= RF_NON_TLS_BSS; 744 745 // We place nobits RelRo sections before plain r/w ones, and non-nobits RelRo 746 // sections after r/w ones, so that the RelRo sections are contiguous. 747 bool IsRelRo = isRelroSection(Sec); 748 if (IsNonTlsNoBits && !IsRelRo) 749 Rank |= RF_NON_TLS_BSS_RO; 750 if (!IsNonTlsNoBits && IsRelRo) 751 Rank |= RF_NON_TLS_BSS_RO; 752 753 // The TLS initialization block needs to be a single contiguous block in a R/W 754 // PT_LOAD, so stick TLS sections directly before the other RelRo R/W 755 // sections. The TLS NOBITS sections are placed here as they don't take up 756 // virtual address space in the PT_LOAD. 757 if (!IsTls) 758 Rank |= RF_NOT_TLS; 759 760 // Within the TLS initialization block, the non-nobits sections need to appear 761 // first. 762 if (IsNoBits) 763 Rank |= RF_BSS; 764 765 // Some architectures have additional ordering restrictions for sections 766 // within the same PT_LOAD. 767 if (Config->EMachine == EM_PPC64) { 768 // PPC64 has a number of special SHT_PROGBITS+SHF_ALLOC+SHF_WRITE sections 769 // that we would like to make sure appear is a specific order to maximize 770 // their coverage by a single signed 16-bit offset from the TOC base 771 // pointer. Conversely, the special .tocbss section should be first among 772 // all SHT_NOBITS sections. This will put it next to the loaded special 773 // PPC64 sections (and, thus, within reach of the TOC base pointer). 774 StringRef Name = Sec->Name; 775 if (Name != ".tocbss") 776 Rank |= RF_PPC_NOT_TOCBSS; 777 778 if (Name == ".opd") 779 Rank |= RF_PPC_OPD; 780 781 if (Name == ".toc1") 782 Rank |= RF_PPC_TOCL; 783 784 if (Name == ".toc") 785 Rank |= RF_PPC_TOC; 786 787 if (Name == ".branch_lt") 788 Rank |= RF_PPC_BRANCH_LT; 789 } 790 if (Config->EMachine == EM_MIPS) { 791 // All sections with SHF_MIPS_GPREL flag should be grouped together 792 // because data in these sections is addressable with a gp relative address. 793 if (Sec->Flags & SHF_MIPS_GPREL) 794 Rank |= RF_MIPS_GPREL; 795 796 if (Sec->Name != ".got") 797 Rank |= RF_MIPS_NOT_GOT; 798 } 799 800 return Rank; 801 } 802 803 static bool compareSections(const BaseCommand *ACmd, const BaseCommand *BCmd) { 804 const OutputSection *A = cast<OutputSection>(ACmd); 805 const OutputSection *B = cast<OutputSection>(BCmd); 806 if (A->SortRank != B->SortRank) 807 return A->SortRank < B->SortRank; 808 if (!(A->SortRank & RF_NOT_ADDR_SET)) 809 return Config->SectionStartMap.lookup(A->Name) < 810 Config->SectionStartMap.lookup(B->Name); 811 return false; 812 } 813 814 void PhdrEntry::add(OutputSection *Sec) { 815 LastSec = Sec; 816 if (!FirstSec) 817 FirstSec = Sec; 818 p_align = std::max(p_align, Sec->Alignment); 819 if (p_type == PT_LOAD) 820 Sec->PtLoad = this; 821 } 822 823 // The beginning and the ending of .rel[a].plt section are marked 824 // with __rel[a]_iplt_{start,end} symbols if it is a statically linked 825 // executable. The runtime needs these symbols in order to resolve 826 // all IRELATIVE relocs on startup. For dynamic executables, we don't 827 // need these symbols, since IRELATIVE relocs are resolved through GOT 828 // and PLT. For details, see http://www.airs.com/blog/archives/403. 829 template <class ELFT> void Writer<ELFT>::addRelIpltSymbols() { 830 if (!Config->Static) 831 return; 832 StringRef S = Config->IsRela ? "__rela_iplt_start" : "__rel_iplt_start"; 833 addOptionalRegular<ELFT>(S, InX::RelaIplt, 0, STV_HIDDEN, STB_WEAK); 834 835 S = Config->IsRela ? "__rela_iplt_end" : "__rel_iplt_end"; 836 addOptionalRegular<ELFT>(S, InX::RelaIplt, -1, STV_HIDDEN, STB_WEAK); 837 } 838 839 template <class ELFT> 840 void Writer<ELFT>::forEachRelSec(std::function<void(InputSectionBase &)> Fn) { 841 // Scan all relocations. Each relocation goes through a series 842 // of tests to determine if it needs special treatment, such as 843 // creating GOT, PLT, copy relocations, etc. 844 // Note that relocations for non-alloc sections are directly 845 // processed by InputSection::relocateNonAlloc. 846 for (InputSectionBase *IS : InputSections) 847 if (IS->Live && isa<InputSection>(IS) && (IS->Flags & SHF_ALLOC)) 848 Fn(*IS); 849 for (EhInputSection *ES : InX::EhFrame->Sections) 850 Fn(*ES); 851 } 852 853 // This function generates assignments for predefined symbols (e.g. _end or 854 // _etext) and inserts them into the commands sequence to be processed at the 855 // appropriate time. This ensures that the value is going to be correct by the 856 // time any references to these symbols are processed and is equivalent to 857 // defining these symbols explicitly in the linker script. 858 template <class ELFT> void Writer<ELFT>::setReservedSymbolSections() { 859 if (ElfSym::GlobalOffsetTable) { 860 // The _GLOBAL_OFFSET_TABLE_ symbol is defined by target convention to 861 // be at some offset from the base of the .got section, usually 0 or the end 862 // of the .got 863 InputSection *GotSection = InX::MipsGot ? cast<InputSection>(InX::MipsGot) 864 : cast<InputSection>(InX::Got); 865 ElfSym::GlobalOffsetTable->Section = GotSection; 866 } 867 868 PhdrEntry *Last = nullptr; 869 PhdrEntry *LastRO = nullptr; 870 871 for (PhdrEntry *P : Phdrs) { 872 if (P->p_type != PT_LOAD) 873 continue; 874 Last = P; 875 if (!(P->p_flags & PF_W)) 876 LastRO = P; 877 } 878 879 if (LastRO) { 880 // _etext is the first location after the last read-only loadable segment. 881 if (ElfSym::Etext1) 882 ElfSym::Etext1->Section = LastRO->LastSec; 883 if (ElfSym::Etext2) 884 ElfSym::Etext2->Section = LastRO->LastSec; 885 } 886 887 if (Last) { 888 // _edata points to the end of the last mapped initialized section. 889 OutputSection *Edata = nullptr; 890 for (OutputSection *OS : OutputSections) { 891 if (OS->Type != SHT_NOBITS) 892 Edata = OS; 893 if (OS == Last->LastSec) 894 break; 895 } 896 897 if (ElfSym::Edata1) 898 ElfSym::Edata1->Section = Edata; 899 if (ElfSym::Edata2) 900 ElfSym::Edata2->Section = Edata; 901 902 // _end is the first location after the uninitialized data region. 903 if (ElfSym::End1) 904 ElfSym::End1->Section = Last->LastSec; 905 if (ElfSym::End2) 906 ElfSym::End2->Section = Last->LastSec; 907 } 908 909 if (ElfSym::Bss) 910 ElfSym::Bss->Section = findSection(".bss"); 911 912 // Setup MIPS _gp_disp/__gnu_local_gp symbols which should 913 // be equal to the _gp symbol's value. 914 if (ElfSym::MipsGp) { 915 // Find GP-relative section with the lowest address 916 // and use this address to calculate default _gp value. 917 for (OutputSection *OS : OutputSections) { 918 if (OS->Flags & SHF_MIPS_GPREL) { 919 ElfSym::MipsGp->Section = OS; 920 ElfSym::MipsGp->Value = 0x7ff0; 921 break; 922 } 923 } 924 } 925 } 926 927 // We want to find how similar two ranks are. 928 // The more branches in getSectionRank that match, the more similar they are. 929 // Since each branch corresponds to a bit flag, we can just use 930 // countLeadingZeros. 931 static int getRankProximityAux(OutputSection *A, OutputSection *B) { 932 return countLeadingZeros(A->SortRank ^ B->SortRank); 933 } 934 935 static int getRankProximity(OutputSection *A, BaseCommand *B) { 936 if (auto *Sec = dyn_cast<OutputSection>(B)) 937 if (Sec->Live) 938 return getRankProximityAux(A, Sec); 939 return -1; 940 } 941 942 // When placing orphan sections, we want to place them after symbol assignments 943 // so that an orphan after 944 // begin_foo = .; 945 // foo : { *(foo) } 946 // end_foo = .; 947 // doesn't break the intended meaning of the begin/end symbols. 948 // We don't want to go over sections since findOrphanPos is the 949 // one in charge of deciding the order of the sections. 950 // We don't want to go over changes to '.', since doing so in 951 // rx_sec : { *(rx_sec) } 952 // . = ALIGN(0x1000); 953 // /* The RW PT_LOAD starts here*/ 954 // rw_sec : { *(rw_sec) } 955 // would mean that the RW PT_LOAD would become unaligned. 956 static bool shouldSkip(BaseCommand *Cmd) { 957 if (isa<OutputSection>(Cmd)) 958 return false; 959 if (auto *Assign = dyn_cast<SymbolAssignment>(Cmd)) 960 return Assign->Name != "."; 961 return true; 962 } 963 964 // We want to place orphan sections so that they share as much 965 // characteristics with their neighbors as possible. For example, if 966 // both are rw, or both are tls. 967 template <typename ELFT> 968 static std::vector<BaseCommand *>::iterator 969 findOrphanPos(std::vector<BaseCommand *>::iterator B, 970 std::vector<BaseCommand *>::iterator E) { 971 OutputSection *Sec = cast<OutputSection>(*E); 972 973 // Find the first element that has as close a rank as possible. 974 auto I = std::max_element(B, E, [=](BaseCommand *A, BaseCommand *B) { 975 return getRankProximity(Sec, A) < getRankProximity(Sec, B); 976 }); 977 if (I == E) 978 return E; 979 980 // Consider all existing sections with the same proximity. 981 int Proximity = getRankProximity(Sec, *I); 982 for (; I != E; ++I) { 983 auto *CurSec = dyn_cast<OutputSection>(*I); 984 if (!CurSec || !CurSec->Live) 985 continue; 986 if (getRankProximity(Sec, CurSec) != Proximity || 987 Sec->SortRank < CurSec->SortRank) 988 break; 989 } 990 991 auto IsLiveSection = [](BaseCommand *Cmd) { 992 auto *OS = dyn_cast<OutputSection>(Cmd); 993 return OS && OS->Live; 994 }; 995 996 auto J = std::find_if(llvm::make_reverse_iterator(I), 997 llvm::make_reverse_iterator(B), IsLiveSection); 998 I = J.base(); 999 1000 // As a special case, if the orphan section is the last section, put 1001 // it at the very end, past any other commands. 1002 // This matches bfd's behavior and is convenient when the linker script fully 1003 // specifies the start of the file, but doesn't care about the end (the non 1004 // alloc sections for example). 1005 auto NextSec = std::find_if(I, E, IsLiveSection); 1006 if (NextSec == E) 1007 return E; 1008 1009 while (I != E && shouldSkip(*I)) 1010 ++I; 1011 return I; 1012 } 1013 1014 // If no layout was provided by linker script, we want to apply default 1015 // sorting for special input sections and handle --symbol-ordering-file. 1016 template <class ELFT> void Writer<ELFT>::sortInputSections() { 1017 assert(!Script->HasSectionsCommand); 1018 1019 // Sort input sections by priority using the list provided 1020 // by --symbol-ordering-file. 1021 DenseMap<SectionBase *, int> Order = buildSectionOrder(); 1022 if (!Order.empty()) 1023 for (BaseCommand *Base : Script->SectionCommands) 1024 if (auto *Sec = dyn_cast<OutputSection>(Base)) 1025 if (Sec->Live) 1026 Sec->sort([&](InputSectionBase *S) { return Order.lookup(S); }); 1027 1028 // Sort input sections by section name suffixes for 1029 // __attribute__((init_priority(N))). 1030 if (OutputSection *Sec = findSection(".init_array")) 1031 Sec->sortInitFini(); 1032 if (OutputSection *Sec = findSection(".fini_array")) 1033 Sec->sortInitFini(); 1034 1035 // Sort input sections by the special rule for .ctors and .dtors. 1036 if (OutputSection *Sec = findSection(".ctors")) 1037 Sec->sortCtorsDtors(); 1038 if (OutputSection *Sec = findSection(".dtors")) 1039 Sec->sortCtorsDtors(); 1040 } 1041 1042 template <class ELFT> void Writer<ELFT>::sortSections() { 1043 Script->adjustSectionsBeforeSorting(); 1044 1045 // Don't sort if using -r. It is not necessary and we want to preserve the 1046 // relative order for SHF_LINK_ORDER sections. 1047 if (Config->Relocatable) 1048 return; 1049 1050 for (BaseCommand *Base : Script->SectionCommands) 1051 if (auto *Sec = dyn_cast<OutputSection>(Base)) 1052 Sec->SortRank = getSectionRank(Sec); 1053 1054 if (!Script->HasSectionsCommand) { 1055 sortInputSections(); 1056 1057 // We know that all the OutputSections are contiguous in this case. 1058 auto E = Script->SectionCommands.end(); 1059 auto I = Script->SectionCommands.begin(); 1060 auto IsSection = [](BaseCommand *Base) { return isa<OutputSection>(Base); }; 1061 I = std::find_if(I, E, IsSection); 1062 E = std::find_if(llvm::make_reverse_iterator(E), 1063 llvm::make_reverse_iterator(I), IsSection) 1064 .base(); 1065 std::stable_sort(I, E, compareSections); 1066 return; 1067 } 1068 1069 // Orphan sections are sections present in the input files which are 1070 // not explicitly placed into the output file by the linker script. 1071 // 1072 // The sections in the linker script are already in the correct 1073 // order. We have to figuere out where to insert the orphan 1074 // sections. 1075 // 1076 // The order of the sections in the script is arbitrary and may not agree with 1077 // compareSections. This means that we cannot easily define a strict weak 1078 // ordering. To see why, consider a comparison of a section in the script and 1079 // one not in the script. We have a two simple options: 1080 // * Make them equivalent (a is not less than b, and b is not less than a). 1081 // The problem is then that equivalence has to be transitive and we can 1082 // have sections a, b and c with only b in a script and a less than c 1083 // which breaks this property. 1084 // * Use compareSectionsNonScript. Given that the script order doesn't have 1085 // to match, we can end up with sections a, b, c, d where b and c are in the 1086 // script and c is compareSectionsNonScript less than b. In which case d 1087 // can be equivalent to c, a to b and d < a. As a concrete example: 1088 // .a (rx) # not in script 1089 // .b (rx) # in script 1090 // .c (ro) # in script 1091 // .d (ro) # not in script 1092 // 1093 // The way we define an order then is: 1094 // * Sort only the orphan sections. They are in the end right now. 1095 // * Move each orphan section to its preferred position. We try 1096 // to put each section in the last position where it it can share 1097 // a PT_LOAD. 1098 // 1099 // There is some ambiguity as to where exactly a new entry should be 1100 // inserted, because Commands contains not only output section 1101 // commands but also other types of commands such as symbol assignment 1102 // expressions. There's no correct answer here due to the lack of the 1103 // formal specification of the linker script. We use heuristics to 1104 // determine whether a new output command should be added before or 1105 // after another commands. For the details, look at shouldSkip 1106 // function. 1107 1108 auto I = Script->SectionCommands.begin(); 1109 auto E = Script->SectionCommands.end(); 1110 auto NonScriptI = std::find_if(I, E, [](BaseCommand *Base) { 1111 if (auto *Sec = dyn_cast<OutputSection>(Base)) 1112 return Sec->Live && Sec->SectionIndex == INT_MAX; 1113 return false; 1114 }); 1115 1116 // Sort the orphan sections. 1117 std::stable_sort(NonScriptI, E, compareSections); 1118 1119 // As a horrible special case, skip the first . assignment if it is before any 1120 // section. We do this because it is common to set a load address by starting 1121 // the script with ". = 0xabcd" and the expectation is that every section is 1122 // after that. 1123 auto FirstSectionOrDotAssignment = 1124 std::find_if(I, E, [](BaseCommand *Cmd) { return !shouldSkip(Cmd); }); 1125 if (FirstSectionOrDotAssignment != E && 1126 isa<SymbolAssignment>(**FirstSectionOrDotAssignment)) 1127 ++FirstSectionOrDotAssignment; 1128 I = FirstSectionOrDotAssignment; 1129 1130 while (NonScriptI != E) { 1131 auto Pos = findOrphanPos<ELFT>(I, NonScriptI); 1132 OutputSection *Orphan = cast<OutputSection>(*NonScriptI); 1133 1134 // As an optimization, find all sections with the same sort rank 1135 // and insert them with one rotate. 1136 unsigned Rank = Orphan->SortRank; 1137 auto End = std::find_if(NonScriptI + 1, E, [=](BaseCommand *Cmd) { 1138 return cast<OutputSection>(Cmd)->SortRank != Rank; 1139 }); 1140 std::rotate(Pos, NonScriptI, End); 1141 NonScriptI = End; 1142 } 1143 1144 Script->adjustSectionsAfterSorting(); 1145 } 1146 1147 static bool compareByFilePosition(InputSection *A, InputSection *B) { 1148 // Synthetic doesn't have link order dependecy, stable_sort will keep it last 1149 if (A->kind() == InputSectionBase::Synthetic || 1150 B->kind() == InputSectionBase::Synthetic) 1151 return false; 1152 InputSection *LA = A->getLinkOrderDep(); 1153 InputSection *LB = B->getLinkOrderDep(); 1154 OutputSection *AOut = LA->getParent(); 1155 OutputSection *BOut = LB->getParent(); 1156 if (AOut != BOut) 1157 return AOut->SectionIndex < BOut->SectionIndex; 1158 return LA->OutSecOff < LB->OutSecOff; 1159 } 1160 1161 // This function is used by the --merge-exidx-entries to detect duplicate 1162 // .ARM.exidx sections. It is Arm only. 1163 // 1164 // The .ARM.exidx section is of the form: 1165 // | PREL31 offset to function | Unwind instructions for function | 1166 // where the unwind instructions are either a small number of unwind 1167 // instructions inlined into the table entry, the special CANT_UNWIND value of 1168 // 0x1 or a PREL31 offset into a .ARM.extab Section that contains unwind 1169 // instructions. 1170 // 1171 // We return true if all the unwind instructions in the .ARM.exidx entries of 1172 // Cur can be merged into the last entry of Prev. 1173 static bool isDuplicateArmExidxSec(InputSection *Prev, InputSection *Cur) { 1174 1175 // References to .ARM.Extab Sections have bit 31 clear and are not the 1176 // special EXIDX_CANTUNWIND bit-pattern. 1177 auto IsExtabRef = [](uint32_t Unwind) { 1178 return (Unwind & 0x80000000) == 0 && Unwind != 0x1; 1179 }; 1180 1181 struct ExidxEntry { 1182 ulittle32_t Fn; 1183 ulittle32_t Unwind; 1184 }; 1185 1186 // Get the last table Entry from the previous .ARM.exidx section. 1187 const ExidxEntry &PrevEntry = *reinterpret_cast<const ExidxEntry *>( 1188 Prev->Data.data() + Prev->getSize() - sizeof(ExidxEntry)); 1189 if (IsExtabRef(PrevEntry.Unwind)) 1190 return false; 1191 1192 // We consider the unwind instructions of an .ARM.exidx table entry 1193 // a duplicate if the previous unwind instructions if: 1194 // - Both are the special EXIDX_CANTUNWIND. 1195 // - Both are the same inline unwind instructions. 1196 // We do not attempt to follow and check links into .ARM.extab tables as 1197 // consecutive identical entries are rare and the effort to check that they 1198 // are identical is high. 1199 1200 if (isa<SyntheticSection>(Cur)) 1201 // Exidx sentinel section has implicit EXIDX_CANTUNWIND; 1202 return PrevEntry.Unwind == 0x1; 1203 1204 ArrayRef<const ExidxEntry> Entries( 1205 reinterpret_cast<const ExidxEntry *>(Cur->Data.data()), 1206 Cur->getSize() / sizeof(ExidxEntry)); 1207 for (const ExidxEntry &Entry : Entries) 1208 if (IsExtabRef(Entry.Unwind) || Entry.Unwind != PrevEntry.Unwind) 1209 return false; 1210 // All table entries in this .ARM.exidx Section can be merged into the 1211 // previous Section. 1212 return true; 1213 } 1214 1215 template <class ELFT> void Writer<ELFT>::resolveShfLinkOrder() { 1216 for (OutputSection *Sec : OutputSections) { 1217 if (!(Sec->Flags & SHF_LINK_ORDER)) 1218 continue; 1219 1220 // Link order may be distributed across several InputSectionDescriptions 1221 // but sort must consider them all at once. 1222 std::vector<InputSection **> ScriptSections; 1223 std::vector<InputSection *> Sections; 1224 for (BaseCommand *Base : Sec->SectionCommands) { 1225 if (auto *ISD = dyn_cast<InputSectionDescription>(Base)) { 1226 for (InputSection *&IS : ISD->Sections) { 1227 ScriptSections.push_back(&IS); 1228 Sections.push_back(IS); 1229 } 1230 } 1231 } 1232 std::stable_sort(Sections.begin(), Sections.end(), compareByFilePosition); 1233 1234 if (Config->MergeArmExidx && !Config->Relocatable && 1235 Config->EMachine == EM_ARM && Sec->Type == SHT_ARM_EXIDX) { 1236 // The EHABI for the Arm Architecture permits consecutive identical 1237 // table entries to be merged. We use a simple implementation that 1238 // removes a .ARM.exidx Input Section if it can be merged into the 1239 // previous one. This does not require any rewriting of InputSection 1240 // contents but misses opportunities for fine grained deduplication where 1241 // only a subset of the InputSection contents can be merged. 1242 int Cur = 1; 1243 int Prev = 0; 1244 int N = Sections.size(); 1245 while (Cur < N) { 1246 if (isDuplicateArmExidxSec(Sections[Prev], Sections[Cur])) 1247 Sections[Cur] = nullptr; 1248 else 1249 Prev = Cur; 1250 ++Cur; 1251 } 1252 } 1253 1254 for (int I = 0, N = Sections.size(); I < N; ++I) 1255 *ScriptSections[I] = Sections[I]; 1256 1257 // Remove the Sections we marked as duplicate earlier. 1258 for (BaseCommand *Base : Sec->SectionCommands) 1259 if (auto *ISD = dyn_cast<InputSectionDescription>(Base)) 1260 ISD->Sections.erase( 1261 std::remove(ISD->Sections.begin(), ISD->Sections.end(), nullptr), 1262 ISD->Sections.end()); 1263 } 1264 } 1265 1266 static void applySynthetic(const std::vector<SyntheticSection *> &Sections, 1267 std::function<void(SyntheticSection *)> Fn) { 1268 for (SyntheticSection *SS : Sections) 1269 if (SS && SS->getParent() && !SS->empty()) 1270 Fn(SS); 1271 } 1272 1273 // In order to allow users to manipulate linker-synthesized sections, 1274 // we had to add synthetic sections to the input section list early, 1275 // even before we make decisions whether they are needed. This allows 1276 // users to write scripts like this: ".mygot : { .got }". 1277 // 1278 // Doing it has an unintended side effects. If it turns out that we 1279 // don't need a .got (for example) at all because there's no 1280 // relocation that needs a .got, we don't want to emit .got. 1281 // 1282 // To deal with the above problem, this function is called after 1283 // scanRelocations is called to remove synthetic sections that turn 1284 // out to be empty. 1285 static void removeUnusedSyntheticSections() { 1286 // All input synthetic sections that can be empty are placed after 1287 // all regular ones. We iterate over them all and exit at first 1288 // non-synthetic. 1289 for (InputSectionBase *S : llvm::reverse(InputSections)) { 1290 SyntheticSection *SS = dyn_cast<SyntheticSection>(S); 1291 if (!SS) 1292 return; 1293 OutputSection *OS = SS->getParent(); 1294 if (!SS->empty() || !OS) 1295 continue; 1296 1297 std::vector<BaseCommand *>::iterator Empty = OS->SectionCommands.end(); 1298 for (auto I = OS->SectionCommands.begin(), E = OS->SectionCommands.end(); 1299 I != E; ++I) { 1300 BaseCommand *B = *I; 1301 if (auto *ISD = dyn_cast<InputSectionDescription>(B)) { 1302 llvm::erase_if(ISD->Sections, 1303 [=](InputSection *IS) { return IS == SS; }); 1304 if (ISD->Sections.empty()) 1305 Empty = I; 1306 } 1307 } 1308 if (Empty != OS->SectionCommands.end()) 1309 OS->SectionCommands.erase(Empty); 1310 1311 // If there are no other sections in the output section, remove it from the 1312 // output. 1313 if (OS->SectionCommands.empty()) 1314 OS->Live = false; 1315 } 1316 } 1317 1318 // Returns true if a symbol can be replaced at load-time by a symbol 1319 // with the same name defined in other ELF executable or DSO. 1320 static bool computeIsPreemptible(const Symbol &B) { 1321 assert(!B.isLocal()); 1322 // Only symbols that appear in dynsym can be preempted. 1323 if (!B.includeInDynsym()) 1324 return false; 1325 1326 // Only default visibility symbols can be preempted. 1327 if (B.Visibility != STV_DEFAULT) 1328 return false; 1329 1330 // At this point copy relocations have not been created yet, so any 1331 // symbol that is not defined locally is preemptible. 1332 if (!B.isDefined()) 1333 return true; 1334 1335 // If we have a dynamic list it specifies which local symbols are preemptible. 1336 if (Config->HasDynamicList) 1337 return false; 1338 1339 if (!Config->Shared) 1340 return false; 1341 1342 // -Bsymbolic means that definitions are not preempted. 1343 if (Config->Bsymbolic || (Config->BsymbolicFunctions && B.isFunc())) 1344 return false; 1345 return true; 1346 } 1347 1348 // Create output section objects and add them to OutputSections. 1349 template <class ELFT> void Writer<ELFT>::finalizeSections() { 1350 Out::DebugInfo = findSection(".debug_info"); 1351 Out::PreinitArray = findSection(".preinit_array"); 1352 Out::InitArray = findSection(".init_array"); 1353 Out::FiniArray = findSection(".fini_array"); 1354 1355 // The linker needs to define SECNAME_start, SECNAME_end and SECNAME_stop 1356 // symbols for sections, so that the runtime can get the start and end 1357 // addresses of each section by section name. Add such symbols. 1358 if (!Config->Relocatable) { 1359 addStartEndSymbols(); 1360 for (BaseCommand *Base : Script->SectionCommands) 1361 if (auto *Sec = dyn_cast<OutputSection>(Base)) 1362 addStartStopSymbols(Sec); 1363 } 1364 1365 // Add _DYNAMIC symbol. Unlike GNU gold, our _DYNAMIC symbol has no type. 1366 // It should be okay as no one seems to care about the type. 1367 // Even the author of gold doesn't remember why gold behaves that way. 1368 // https://sourceware.org/ml/binutils/2002-03/msg00360.html 1369 if (InX::DynSymTab) 1370 Symtab->addRegular<ELFT>("_DYNAMIC", STV_HIDDEN, STT_NOTYPE, 0 /*Value*/, 1371 /*Size=*/0, STB_WEAK, InX::Dynamic, 1372 /*File=*/nullptr); 1373 1374 // Define __rel[a]_iplt_{start,end} symbols if needed. 1375 addRelIpltSymbols(); 1376 1377 // This responsible for splitting up .eh_frame section into 1378 // pieces. The relocation scan uses those pieces, so this has to be 1379 // earlier. 1380 applySynthetic({InX::EhFrame}, 1381 [](SyntheticSection *SS) { SS->finalizeContents(); }); 1382 1383 for (Symbol *S : Symtab->getSymbols()) 1384 S->IsPreemptible |= computeIsPreemptible(*S); 1385 1386 // Scan relocations. This must be done after every symbol is declared so that 1387 // we can correctly decide if a dynamic relocation is needed. 1388 if (!Config->Relocatable) 1389 forEachRelSec(scanRelocations<ELFT>); 1390 1391 if (InX::Plt && !InX::Plt->empty()) 1392 InX::Plt->addSymbols(); 1393 if (InX::Iplt && !InX::Iplt->empty()) 1394 InX::Iplt->addSymbols(); 1395 1396 // Now that we have defined all possible global symbols including linker- 1397 // synthesized ones. Visit all symbols to give the finishing touches. 1398 for (Symbol *Sym : Symtab->getSymbols()) { 1399 if (!includeInSymtab(*Sym)) 1400 continue; 1401 if (InX::SymTab) 1402 InX::SymTab->addSymbol(Sym); 1403 1404 if (InX::DynSymTab && Sym->includeInDynsym()) { 1405 InX::DynSymTab->addSymbol(Sym); 1406 if (auto *SS = dyn_cast<SharedSymbol>(Sym)) 1407 if (cast<SharedFile<ELFT>>(Sym->File)->IsNeeded) 1408 In<ELFT>::VerNeed->addSymbol(SS); 1409 } 1410 } 1411 1412 // Do not proceed if there was an undefined symbol. 1413 if (errorCount()) 1414 return; 1415 1416 addPredefinedSections(); 1417 removeUnusedSyntheticSections(); 1418 1419 sortSections(); 1420 Script->removeEmptyCommands(); 1421 1422 // Now that we have the final list, create a list of all the 1423 // OutputSections for convenience. 1424 for (BaseCommand *Base : Script->SectionCommands) 1425 if (auto *Sec = dyn_cast<OutputSection>(Base)) 1426 OutputSections.push_back(Sec); 1427 1428 // Prefer command line supplied address over other constraints. 1429 for (OutputSection *Sec : OutputSections) { 1430 auto I = Config->SectionStartMap.find(Sec->Name); 1431 if (I != Config->SectionStartMap.end()) 1432 Sec->AddrExpr = [=] { return I->second; }; 1433 } 1434 1435 // This is a bit of a hack. A value of 0 means undef, so we set it 1436 // to 1 t make __ehdr_start defined. The section number is not 1437 // particularly relevant. 1438 Out::ElfHeader->SectionIndex = 1; 1439 1440 unsigned I = 1; 1441 for (OutputSection *Sec : OutputSections) { 1442 Sec->SectionIndex = I++; 1443 Sec->ShName = InX::ShStrTab->addString(Sec->Name); 1444 } 1445 1446 // Binary and relocatable output does not have PHDRS. 1447 // The headers have to be created before finalize as that can influence the 1448 // image base and the dynamic section on mips includes the image base. 1449 if (!Config->Relocatable && !Config->OFormatBinary) { 1450 Phdrs = Script->hasPhdrsCommands() ? Script->createPhdrs() : createPhdrs(); 1451 addPtArmExid(Phdrs); 1452 Out::ProgramHeaders->Size = sizeof(Elf_Phdr) * Phdrs.size(); 1453 } 1454 1455 // Some symbols are defined in term of program headers. Now that we 1456 // have the headers, we can find out which sections they point to. 1457 setReservedSymbolSections(); 1458 1459 // Dynamic section must be the last one in this list and dynamic 1460 // symbol table section (DynSymTab) must be the first one. 1461 applySynthetic( 1462 {InX::DynSymTab, InX::Bss, InX::BssRelRo, InX::GnuHashTab, 1463 InX::HashTab, InX::SymTab, InX::ShStrTab, InX::StrTab, 1464 In<ELFT>::VerDef, InX::DynStrTab, InX::Got, InX::MipsGot, 1465 InX::IgotPlt, InX::GotPlt, InX::RelaDyn, InX::RelaIplt, 1466 InX::RelaPlt, InX::Plt, InX::Iplt, InX::EhFrameHdr, 1467 In<ELFT>::VerSym, In<ELFT>::VerNeed, InX::Dynamic}, 1468 [](SyntheticSection *SS) { SS->finalizeContents(); }); 1469 1470 if (!Script->HasSectionsCommand && !Config->Relocatable) 1471 fixSectionAlignments(); 1472 1473 // After link order processing .ARM.exidx sections can be deduplicated, which 1474 // needs to be resolved before any other address dependent operation. 1475 resolveShfLinkOrder(); 1476 1477 // Some architectures need to generate content that depends on the address 1478 // of InputSections. For example some architectures use small displacements 1479 // for jump instructions that is is the linker's responsibility for creating 1480 // range extension thunks for. As the generation of the content may also 1481 // alter InputSection addresses we must converge to a fixed point. 1482 if (Target->NeedsThunks || Config->AndroidPackDynRelocs) { 1483 ThunkCreator TC; 1484 AArch64Err843419Patcher A64P; 1485 bool Changed; 1486 do { 1487 Script->assignAddresses(); 1488 Changed = false; 1489 if (Target->NeedsThunks) 1490 Changed |= TC.createThunks(OutputSections); 1491 if (Config->FixCortexA53Errata843419) { 1492 if (Changed) 1493 Script->assignAddresses(); 1494 Changed |= A64P.createFixes(); 1495 } 1496 if (InX::MipsGot) 1497 InX::MipsGot->updateAllocSize(); 1498 Changed |= InX::RelaDyn->updateAllocSize(); 1499 } while (Changed); 1500 } 1501 1502 // Fill other section headers. The dynamic table is finalized 1503 // at the end because some tags like RELSZ depend on result 1504 // of finalizing other sections. 1505 for (OutputSection *Sec : OutputSections) 1506 Sec->finalize<ELFT>(); 1507 1508 // createThunks may have added local symbols to the static symbol table 1509 applySynthetic({InX::SymTab}, 1510 [](SyntheticSection *SS) { SS->postThunkContents(); }); 1511 } 1512 1513 template <class ELFT> void Writer<ELFT>::addPredefinedSections() { 1514 // ARM ABI requires .ARM.exidx to be terminated by some piece of data. 1515 // We have the terminater synthetic section class. Add that at the end. 1516 OutputSection *Cmd = findSection(".ARM.exidx"); 1517 if (!Cmd || !Cmd->Live || Config->Relocatable) 1518 return; 1519 1520 auto *Sentinel = make<ARMExidxSentinelSection>(); 1521 Cmd->addSection(Sentinel); 1522 } 1523 1524 // The linker is expected to define SECNAME_start and SECNAME_end 1525 // symbols for a few sections. This function defines them. 1526 template <class ELFT> void Writer<ELFT>::addStartEndSymbols() { 1527 auto Define = [&](StringRef Start, StringRef End, OutputSection *OS) { 1528 // These symbols resolve to the image base if the section does not exist. 1529 // A special value -1 indicates end of the section. 1530 if (OS) { 1531 addOptionalRegular<ELFT>(Start, OS, 0); 1532 addOptionalRegular<ELFT>(End, OS, -1); 1533 } else { 1534 if (Config->Pic) 1535 OS = Out::ElfHeader; 1536 addOptionalRegular<ELFT>(Start, OS, 0); 1537 addOptionalRegular<ELFT>(End, OS, 0); 1538 } 1539 }; 1540 1541 Define("__preinit_array_start", "__preinit_array_end", Out::PreinitArray); 1542 Define("__init_array_start", "__init_array_end", Out::InitArray); 1543 Define("__fini_array_start", "__fini_array_end", Out::FiniArray); 1544 1545 if (OutputSection *Sec = findSection(".ARM.exidx")) 1546 Define("__exidx_start", "__exidx_end", Sec); 1547 } 1548 1549 // If a section name is valid as a C identifier (which is rare because of 1550 // the leading '.'), linkers are expected to define __start_<secname> and 1551 // __stop_<secname> symbols. They are at beginning and end of the section, 1552 // respectively. This is not requested by the ELF standard, but GNU ld and 1553 // gold provide the feature, and used by many programs. 1554 template <class ELFT> 1555 void Writer<ELFT>::addStartStopSymbols(OutputSection *Sec) { 1556 StringRef S = Sec->Name; 1557 if (!isValidCIdentifier(S)) 1558 return; 1559 addOptionalRegular<ELFT>(Saver.save("__start_" + S), Sec, 0, STV_DEFAULT); 1560 addOptionalRegular<ELFT>(Saver.save("__stop_" + S), Sec, -1, STV_DEFAULT); 1561 } 1562 1563 static bool needsPtLoad(OutputSection *Sec) { 1564 if (!(Sec->Flags & SHF_ALLOC)) 1565 return false; 1566 1567 // Don't allocate VA space for TLS NOBITS sections. The PT_TLS PHDR is 1568 // responsible for allocating space for them, not the PT_LOAD that 1569 // contains the TLS initialization image. 1570 if (Sec->Flags & SHF_TLS && Sec->Type == SHT_NOBITS) 1571 return false; 1572 return true; 1573 } 1574 1575 // Linker scripts are responsible for aligning addresses. Unfortunately, most 1576 // linker scripts are designed for creating two PT_LOADs only, one RX and one 1577 // RW. This means that there is no alignment in the RO to RX transition and we 1578 // cannot create a PT_LOAD there. 1579 static uint64_t computeFlags(uint64_t Flags) { 1580 if (Config->Omagic) 1581 return PF_R | PF_W | PF_X; 1582 if (Config->SingleRoRx && !(Flags & PF_W)) 1583 return Flags | PF_X; 1584 return Flags; 1585 } 1586 1587 // Decide which program headers to create and which sections to include in each 1588 // one. 1589 template <class ELFT> std::vector<PhdrEntry *> Writer<ELFT>::createPhdrs() { 1590 std::vector<PhdrEntry *> Ret; 1591 auto AddHdr = [&](unsigned Type, unsigned Flags) -> PhdrEntry * { 1592 Ret.push_back(make<PhdrEntry>(Type, Flags)); 1593 return Ret.back(); 1594 }; 1595 1596 // The first phdr entry is PT_PHDR which describes the program header itself. 1597 AddHdr(PT_PHDR, PF_R)->add(Out::ProgramHeaders); 1598 1599 // PT_INTERP must be the second entry if exists. 1600 if (OutputSection *Cmd = findSection(".interp")) 1601 AddHdr(PT_INTERP, Cmd->getPhdrFlags())->add(Cmd); 1602 1603 // Add the first PT_LOAD segment for regular output sections. 1604 uint64_t Flags = computeFlags(PF_R); 1605 PhdrEntry *Load = AddHdr(PT_LOAD, Flags); 1606 1607 // Add the headers. We will remove them if they don't fit. 1608 Load->add(Out::ElfHeader); 1609 Load->add(Out::ProgramHeaders); 1610 1611 for (OutputSection *Sec : OutputSections) { 1612 if (!(Sec->Flags & SHF_ALLOC)) 1613 break; 1614 if (!needsPtLoad(Sec)) 1615 continue; 1616 1617 // Segments are contiguous memory regions that has the same attributes 1618 // (e.g. executable or writable). There is one phdr for each segment. 1619 // Therefore, we need to create a new phdr when the next section has 1620 // different flags or is loaded at a discontiguous address using AT linker 1621 // script command. 1622 uint64_t NewFlags = computeFlags(Sec->getPhdrFlags()); 1623 if (Sec->LMAExpr || Flags != NewFlags) { 1624 Load = AddHdr(PT_LOAD, NewFlags); 1625 Flags = NewFlags; 1626 } 1627 1628 Load->add(Sec); 1629 } 1630 1631 // Add a TLS segment if any. 1632 PhdrEntry *TlsHdr = make<PhdrEntry>(PT_TLS, PF_R); 1633 for (OutputSection *Sec : OutputSections) 1634 if (Sec->Flags & SHF_TLS) 1635 TlsHdr->add(Sec); 1636 if (TlsHdr->FirstSec) 1637 Ret.push_back(TlsHdr); 1638 1639 // Add an entry for .dynamic. 1640 if (InX::DynSymTab) 1641 AddHdr(PT_DYNAMIC, InX::Dynamic->getParent()->getPhdrFlags()) 1642 ->add(InX::Dynamic->getParent()); 1643 1644 // PT_GNU_RELRO includes all sections that should be marked as 1645 // read-only by dynamic linker after proccessing relocations. 1646 // Current dynamic loaders only support one PT_GNU_RELRO PHDR, give 1647 // an error message if more than one PT_GNU_RELRO PHDR is required. 1648 PhdrEntry *RelRo = make<PhdrEntry>(PT_GNU_RELRO, PF_R); 1649 bool InRelroPhdr = false; 1650 bool IsRelroFinished = false; 1651 for (OutputSection *Sec : OutputSections) { 1652 if (!needsPtLoad(Sec)) 1653 continue; 1654 if (isRelroSection(Sec)) { 1655 InRelroPhdr = true; 1656 if (!IsRelroFinished) 1657 RelRo->add(Sec); 1658 else 1659 error("section: " + Sec->Name + " is not contiguous with other relro" + 1660 " sections"); 1661 } else if (InRelroPhdr) { 1662 InRelroPhdr = false; 1663 IsRelroFinished = true; 1664 } 1665 } 1666 if (RelRo->FirstSec) 1667 Ret.push_back(RelRo); 1668 1669 // PT_GNU_EH_FRAME is a special section pointing on .eh_frame_hdr. 1670 if (!InX::EhFrame->empty() && InX::EhFrameHdr && InX::EhFrame->getParent() && 1671 InX::EhFrameHdr->getParent()) 1672 AddHdr(PT_GNU_EH_FRAME, InX::EhFrameHdr->getParent()->getPhdrFlags()) 1673 ->add(InX::EhFrameHdr->getParent()); 1674 1675 // PT_OPENBSD_RANDOMIZE is an OpenBSD-specific feature. That makes 1676 // the dynamic linker fill the segment with random data. 1677 if (OutputSection *Cmd = findSection(".openbsd.randomdata")) 1678 AddHdr(PT_OPENBSD_RANDOMIZE, Cmd->getPhdrFlags())->add(Cmd); 1679 1680 // PT_GNU_STACK is a special section to tell the loader to make the 1681 // pages for the stack non-executable. If you really want an executable 1682 // stack, you can pass -z execstack, but that's not recommended for 1683 // security reasons. 1684 unsigned Perm; 1685 if (Config->ZExecstack) 1686 Perm = PF_R | PF_W | PF_X; 1687 else 1688 Perm = PF_R | PF_W; 1689 AddHdr(PT_GNU_STACK, Perm)->p_memsz = Config->ZStackSize; 1690 1691 // PT_OPENBSD_WXNEEDED is a OpenBSD-specific header to mark the executable 1692 // is expected to perform W^X violations, such as calling mprotect(2) or 1693 // mmap(2) with PROT_WRITE | PROT_EXEC, which is prohibited by default on 1694 // OpenBSD. 1695 if (Config->ZWxneeded) 1696 AddHdr(PT_OPENBSD_WXNEEDED, PF_X); 1697 1698 // Create one PT_NOTE per a group of contiguous .note sections. 1699 PhdrEntry *Note = nullptr; 1700 for (OutputSection *Sec : OutputSections) { 1701 if (Sec->Type == SHT_NOTE) { 1702 if (!Note || Sec->LMAExpr) 1703 Note = AddHdr(PT_NOTE, PF_R); 1704 Note->add(Sec); 1705 } else { 1706 Note = nullptr; 1707 } 1708 } 1709 return Ret; 1710 } 1711 1712 template <class ELFT> 1713 void Writer<ELFT>::addPtArmExid(std::vector<PhdrEntry *> &Phdrs) { 1714 if (Config->EMachine != EM_ARM) 1715 return; 1716 auto I = llvm::find_if(OutputSections, [](OutputSection *Cmd) { 1717 return Cmd->Type == SHT_ARM_EXIDX; 1718 }); 1719 if (I == OutputSections.end()) 1720 return; 1721 1722 // PT_ARM_EXIDX is the ARM EHABI equivalent of PT_GNU_EH_FRAME 1723 PhdrEntry *ARMExidx = make<PhdrEntry>(PT_ARM_EXIDX, PF_R); 1724 ARMExidx->add(*I); 1725 Phdrs.push_back(ARMExidx); 1726 } 1727 1728 // The first section of each PT_LOAD, the first section in PT_GNU_RELRO and the 1729 // first section after PT_GNU_RELRO have to be page aligned so that the dynamic 1730 // linker can set the permissions. 1731 template <class ELFT> void Writer<ELFT>::fixSectionAlignments() { 1732 auto PageAlign = [](OutputSection *Cmd) { 1733 if (Cmd && !Cmd->AddrExpr) 1734 Cmd->AddrExpr = [=] { 1735 return alignTo(Script->getDot(), Config->MaxPageSize); 1736 }; 1737 }; 1738 1739 for (const PhdrEntry *P : Phdrs) 1740 if (P->p_type == PT_LOAD && P->FirstSec) 1741 PageAlign(P->FirstSec); 1742 1743 for (const PhdrEntry *P : Phdrs) { 1744 if (P->p_type != PT_GNU_RELRO) 1745 continue; 1746 if (P->FirstSec) 1747 PageAlign(P->FirstSec); 1748 // Find the first section after PT_GNU_RELRO. If it is in a PT_LOAD we 1749 // have to align it to a page. 1750 auto End = OutputSections.end(); 1751 auto I = std::find(OutputSections.begin(), End, P->LastSec); 1752 if (I == End || (I + 1) == End) 1753 continue; 1754 OutputSection *Cmd = (*(I + 1)); 1755 if (needsPtLoad(Cmd)) 1756 PageAlign(Cmd); 1757 } 1758 } 1759 1760 // Adjusts the file alignment for a given output section and returns 1761 // its new file offset. The file offset must be the same with its 1762 // virtual address (modulo the page size) so that the loader can load 1763 // executables without any address adjustment. 1764 static uint64_t getFileAlignment(uint64_t Off, OutputSection *Cmd) { 1765 // If the section is not in a PT_LOAD, we just have to align it. 1766 if (!Cmd->PtLoad) 1767 return alignTo(Off, Cmd->Alignment); 1768 1769 OutputSection *First = Cmd->PtLoad->FirstSec; 1770 // The first section in a PT_LOAD has to have congruent offset and address 1771 // module the page size. 1772 if (Cmd == First) 1773 return alignTo(Off, std::max<uint64_t>(Cmd->Alignment, Config->MaxPageSize), 1774 Cmd->Addr); 1775 1776 // If two sections share the same PT_LOAD the file offset is calculated 1777 // using this formula: Off2 = Off1 + (VA2 - VA1). 1778 return First->Offset + Cmd->Addr - First->Addr; 1779 } 1780 1781 static uint64_t setOffset(OutputSection *Cmd, uint64_t Off) { 1782 if (Cmd->Type == SHT_NOBITS) { 1783 Cmd->Offset = Off; 1784 return Off; 1785 } 1786 1787 Off = getFileAlignment(Off, Cmd); 1788 Cmd->Offset = Off; 1789 return Off + Cmd->Size; 1790 } 1791 1792 template <class ELFT> void Writer<ELFT>::assignFileOffsetsBinary() { 1793 uint64_t Off = 0; 1794 for (OutputSection *Sec : OutputSections) 1795 if (Sec->Flags & SHF_ALLOC) 1796 Off = setOffset(Sec, Off); 1797 FileSize = alignTo(Off, Config->Wordsize); 1798 } 1799 1800 // Assign file offsets to output sections. 1801 template <class ELFT> void Writer<ELFT>::assignFileOffsets() { 1802 uint64_t Off = 0; 1803 Off = setOffset(Out::ElfHeader, Off); 1804 Off = setOffset(Out::ProgramHeaders, Off); 1805 1806 PhdrEntry *LastRX = nullptr; 1807 for (PhdrEntry *P : Phdrs) 1808 if (P->p_type == PT_LOAD && (P->p_flags & PF_X)) 1809 LastRX = P; 1810 1811 for (OutputSection *Sec : OutputSections) { 1812 Off = setOffset(Sec, Off); 1813 if (Script->HasSectionsCommand) 1814 continue; 1815 // If this is a last section of the last executable segment and that 1816 // segment is the last loadable segment, align the offset of the 1817 // following section to avoid loading non-segments parts of the file. 1818 if (LastRX && LastRX->LastSec == Sec) 1819 Off = alignTo(Off, Target->PageSize); 1820 } 1821 1822 SectionHeaderOff = alignTo(Off, Config->Wordsize); 1823 FileSize = SectionHeaderOff + (OutputSections.size() + 1) * sizeof(Elf_Shdr); 1824 } 1825 1826 // Finalize the program headers. We call this function after we assign 1827 // file offsets and VAs to all sections. 1828 template <class ELFT> void Writer<ELFT>::setPhdrs() { 1829 for (PhdrEntry *P : Phdrs) { 1830 OutputSection *First = P->FirstSec; 1831 OutputSection *Last = P->LastSec; 1832 if (First) { 1833 P->p_filesz = Last->Offset - First->Offset; 1834 if (Last->Type != SHT_NOBITS) 1835 P->p_filesz += Last->Size; 1836 P->p_memsz = Last->Addr + Last->Size - First->Addr; 1837 P->p_offset = First->Offset; 1838 P->p_vaddr = First->Addr; 1839 if (!P->HasLMA) 1840 P->p_paddr = First->getLMA(); 1841 } 1842 if (P->p_type == PT_LOAD) 1843 P->p_align = std::max<uint64_t>(P->p_align, Config->MaxPageSize); 1844 else if (P->p_type == PT_GNU_RELRO) { 1845 P->p_align = 1; 1846 // The glibc dynamic loader rounds the size down, so we need to round up 1847 // to protect the last page. This is a no-op on FreeBSD which always 1848 // rounds up. 1849 P->p_memsz = alignTo(P->p_memsz, Target->PageSize); 1850 } 1851 1852 // The TLS pointer goes after PT_TLS. At least glibc will align it, 1853 // so round up the size to make sure the offsets are correct. 1854 if (P->p_type == PT_TLS) { 1855 Out::TlsPhdr = P; 1856 if (P->p_memsz) 1857 P->p_memsz = alignTo(P->p_memsz, P->p_align); 1858 } 1859 } 1860 } 1861 1862 // The entry point address is chosen in the following ways. 1863 // 1864 // 1. the '-e' entry command-line option; 1865 // 2. the ENTRY(symbol) command in a linker control script; 1866 // 3. the value of the symbol _start, if present; 1867 // 4. the number represented by the entry symbol, if it is a number; 1868 // 5. the address of the first byte of the .text section, if present; 1869 // 6. the address 0. 1870 template <class ELFT> uint64_t Writer<ELFT>::getEntryAddr() { 1871 // Case 1, 2 or 3 1872 if (Symbol *B = Symtab->find(Config->Entry)) 1873 return B->getVA(); 1874 1875 // Case 4 1876 uint64_t Addr; 1877 if (to_integer(Config->Entry, Addr)) 1878 return Addr; 1879 1880 // Case 5 1881 if (OutputSection *Sec = findSection(".text")) { 1882 if (Config->WarnMissingEntry) 1883 warn("cannot find entry symbol " + Config->Entry + "; defaulting to 0x" + 1884 utohexstr(Sec->Addr)); 1885 return Sec->Addr; 1886 } 1887 1888 // Case 6 1889 if (Config->WarnMissingEntry) 1890 warn("cannot find entry symbol " + Config->Entry + 1891 "; not setting start address"); 1892 return 0; 1893 } 1894 1895 static uint16_t getELFType() { 1896 if (Config->Pic) 1897 return ET_DYN; 1898 if (Config->Relocatable) 1899 return ET_REL; 1900 return ET_EXEC; 1901 } 1902 1903 template <class ELFT> void Writer<ELFT>::writeHeader() { 1904 uint8_t *Buf = Buffer->getBufferStart(); 1905 memcpy(Buf, "\177ELF", 4); 1906 1907 // Write the ELF header. 1908 auto *EHdr = reinterpret_cast<Elf_Ehdr *>(Buf); 1909 EHdr->e_ident[EI_CLASS] = Config->Is64 ? ELFCLASS64 : ELFCLASS32; 1910 EHdr->e_ident[EI_DATA] = Config->IsLE ? ELFDATA2LSB : ELFDATA2MSB; 1911 EHdr->e_ident[EI_VERSION] = EV_CURRENT; 1912 EHdr->e_ident[EI_OSABI] = Config->OSABI; 1913 EHdr->e_type = getELFType(); 1914 EHdr->e_machine = Config->EMachine; 1915 EHdr->e_version = EV_CURRENT; 1916 EHdr->e_entry = getEntryAddr(); 1917 EHdr->e_shoff = SectionHeaderOff; 1918 EHdr->e_flags = Config->EFlags; 1919 EHdr->e_ehsize = sizeof(Elf_Ehdr); 1920 EHdr->e_phnum = Phdrs.size(); 1921 EHdr->e_shentsize = sizeof(Elf_Shdr); 1922 EHdr->e_shnum = OutputSections.size() + 1; 1923 EHdr->e_shstrndx = InX::ShStrTab->getParent()->SectionIndex; 1924 1925 if (!Config->Relocatable) { 1926 EHdr->e_phoff = sizeof(Elf_Ehdr); 1927 EHdr->e_phentsize = sizeof(Elf_Phdr); 1928 } 1929 1930 // Write the program header table. 1931 auto *HBuf = reinterpret_cast<Elf_Phdr *>(Buf + EHdr->e_phoff); 1932 for (PhdrEntry *P : Phdrs) { 1933 HBuf->p_type = P->p_type; 1934 HBuf->p_flags = P->p_flags; 1935 HBuf->p_offset = P->p_offset; 1936 HBuf->p_vaddr = P->p_vaddr; 1937 HBuf->p_paddr = P->p_paddr; 1938 HBuf->p_filesz = P->p_filesz; 1939 HBuf->p_memsz = P->p_memsz; 1940 HBuf->p_align = P->p_align; 1941 ++HBuf; 1942 } 1943 1944 // Write the section header table. Note that the first table entry is null. 1945 auto *SHdrs = reinterpret_cast<Elf_Shdr *>(Buf + EHdr->e_shoff); 1946 for (OutputSection *Sec : OutputSections) 1947 Sec->writeHeaderTo<ELFT>(++SHdrs); 1948 } 1949 1950 // Open a result file. 1951 template <class ELFT> void Writer<ELFT>::openFile() { 1952 if (!Config->Is64 && FileSize > UINT32_MAX) { 1953 error("output file too large: " + Twine(FileSize) + " bytes"); 1954 return; 1955 } 1956 1957 unlinkAsync(Config->OutputFile); 1958 unsigned Flags = 0; 1959 if (!Config->Relocatable) 1960 Flags = FileOutputBuffer::F_executable; 1961 Expected<std::unique_ptr<FileOutputBuffer>> BufferOrErr = 1962 FileOutputBuffer::create(Config->OutputFile, FileSize, Flags); 1963 1964 if (!BufferOrErr) 1965 error("failed to open " + Config->OutputFile + ": " + 1966 llvm::toString(BufferOrErr.takeError())); 1967 else 1968 Buffer = std::move(*BufferOrErr); 1969 } 1970 1971 template <class ELFT> void Writer<ELFT>::writeSectionsBinary() { 1972 uint8_t *Buf = Buffer->getBufferStart(); 1973 for (OutputSection *Sec : OutputSections) 1974 if (Sec->Flags & SHF_ALLOC) 1975 Sec->writeTo<ELFT>(Buf + Sec->Offset); 1976 } 1977 1978 static void fillTrap(uint8_t *I, uint8_t *End) { 1979 for (; I + 4 <= End; I += 4) 1980 memcpy(I, &Target->TrapInstr, 4); 1981 } 1982 1983 // Fill the last page of executable segments with trap instructions 1984 // instead of leaving them as zero. Even though it is not required by any 1985 // standard, it is in general a good thing to do for security reasons. 1986 // 1987 // We'll leave other pages in segments as-is because the rest will be 1988 // overwritten by output sections. 1989 template <class ELFT> void Writer<ELFT>::writeTrapInstr() { 1990 if (Script->HasSectionsCommand) 1991 return; 1992 1993 // Fill the last page. 1994 uint8_t *Buf = Buffer->getBufferStart(); 1995 for (PhdrEntry *P : Phdrs) 1996 if (P->p_type == PT_LOAD && (P->p_flags & PF_X)) 1997 fillTrap(Buf + alignDown(P->p_offset + P->p_filesz, Target->PageSize), 1998 Buf + alignTo(P->p_offset + P->p_filesz, Target->PageSize)); 1999 2000 // Round up the file size of the last segment to the page boundary iff it is 2001 // an executable segment to ensure that other tools don't accidentally 2002 // trim the instruction padding (e.g. when stripping the file). 2003 PhdrEntry *Last = nullptr; 2004 for (PhdrEntry *P : Phdrs) 2005 if (P->p_type == PT_LOAD) 2006 Last = P; 2007 2008 if (Last && (Last->p_flags & PF_X)) 2009 Last->p_memsz = Last->p_filesz = alignTo(Last->p_filesz, Target->PageSize); 2010 } 2011 2012 // Write section contents to a mmap'ed file. 2013 template <class ELFT> void Writer<ELFT>::writeSections() { 2014 uint8_t *Buf = Buffer->getBufferStart(); 2015 2016 // PPC64 needs to process relocations in the .opd section 2017 // before processing relocations in code-containing sections. 2018 if (auto *OpdCmd = findSection(".opd")) { 2019 Out::Opd = OpdCmd; 2020 Out::OpdBuf = Buf + Out::Opd->Offset; 2021 OpdCmd->template writeTo<ELFT>(Buf + Out::Opd->Offset); 2022 } 2023 2024 OutputSection *EhFrameHdr = nullptr; 2025 if (InX::EhFrameHdr && !InX::EhFrameHdr->empty()) 2026 EhFrameHdr = InX::EhFrameHdr->getParent(); 2027 2028 // In -r or -emit-relocs mode, write the relocation sections first as in 2029 // ELf_Rel targets we might find out that we need to modify the relocated 2030 // section while doing it. 2031 for (OutputSection *Sec : OutputSections) 2032 if (Sec->Type == SHT_REL || Sec->Type == SHT_RELA) 2033 Sec->writeTo<ELFT>(Buf + Sec->Offset); 2034 2035 for (OutputSection *Sec : OutputSections) 2036 if (Sec != Out::Opd && Sec != EhFrameHdr && Sec->Type != SHT_REL && 2037 Sec->Type != SHT_RELA) 2038 Sec->writeTo<ELFT>(Buf + Sec->Offset); 2039 2040 // The .eh_frame_hdr depends on .eh_frame section contents, therefore 2041 // it should be written after .eh_frame is written. 2042 if (EhFrameHdr) 2043 EhFrameHdr->writeTo<ELFT>(Buf + EhFrameHdr->Offset); 2044 } 2045 2046 template <class ELFT> void Writer<ELFT>::writeBuildId() { 2047 if (!InX::BuildId || !InX::BuildId->getParent()) 2048 return; 2049 2050 // Compute a hash of all sections of the output file. 2051 uint8_t *Start = Buffer->getBufferStart(); 2052 uint8_t *End = Start + FileSize; 2053 InX::BuildId->writeBuildId({Start, End}); 2054 } 2055 2056 template void elf::writeResult<ELF32LE>(); 2057 template void elf::writeResult<ELF32BE>(); 2058 template void elf::writeResult<ELF64LE>(); 2059 template void elf::writeResult<ELF64BE>(); 2060 2061 template void elf::addReservedSymbols<ELF32LE>(); 2062 template void elf::addReservedSymbols<ELF32BE>(); 2063 template void elf::addReservedSymbols<ELF64LE>(); 2064 template void elf::addReservedSymbols<ELF64BE>(); 2065